Graphical Methods for Defense Against False-data Injection Attacks on Power System State Estimation

TL;DR

This paper introduces graphical methods to defend power system state estimation against false-data injection attacks by optimally securing measurements, using Steiner tree-based algorithms to minimize protection costs while maintaining security.

Contribution

It formulates the optimal measurement protection problem as a Steiner tree problem and proposes both exact and approximation algorithms for efficient defense.

Findings

The proposed algorithms effectively secure power system state estimation.

Approximation algorithms significantly reduce computational complexity.

The methods are validated on IEEE power system test cases.

Abstract

The normal operation of power system relies on accurate state estimation that faithfully reflects the physical aspects of the electrical power grids. However, recent research shows that carefully synthesized false-data injection attacks can bypass the security system and introduce arbitrary errors to state estimates. In this paper, we use graphical methods to study defending mechanisms against false-data injection attacks on power system state estimation. By securing carefully selected meter measurements, no false data injection attack can be launched to compromise any set of state variables. We characterize the optimal protection problem, which protects the state variables with minimum number of measurements, as a variant Steiner tree problem in a graph. Based on the graphical characterization, we propose both exact and reduced-complexity approximation algorithms. In particular, we show that the proposed tree-pruning based approximation algorithm significantly reduces computational complexity, while yielding negligible performance degradation compared with the optimal algorithms. The advantageous performance of the proposed defending mechanisms is verified in IEEE standard power system testcases.